Blade health state monitoring method and device, medium and equipment
By constructing a blade health status dataset and combining supervised and unsupervised learning models, real-time monitoring of wind turbine blades was achieved, solving the problems of insufficient real-time performance and low accuracy in existing technologies, and improving monitoring efficiency and accuracy.
Patent Information
- Application Number
- CN202411000164.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2026-01-27
AI Technical Summary
Existing methods for monitoring wind turbine blades suffer from insufficient real-time performance, low efficiency, and poor accuracy, making it impossible to achieve efficient health status monitoring.
By constructing a leaf health status dataset, combining supervised and unsupervised learning models for periodic updates, utilizing supervised models for real-time monitoring, and combining data collection with fiber optic sensors, real-time monitoring of leaf health status can be achieved.
It enables real-time monitoring of leaf health status, improves monitoring efficiency and accuracy, reduces the need for manual inspection, and has the capability of second-level threshold alarm.
Smart Images
Figure CN121408151A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power distribution equipment testing, specifically to a method, device, medium, and equipment for monitoring the health status of blades. Background Technology
[0002] With increasing awareness of environmental protection and sustainable development, wind power has gradually become a focus of attention. As a clean energy source, wind power can not only reduce dependence on fossil fuels but also reduce environmental pollution. One of the core components of a wind turbine is the turbine blade, whose performance directly affects the power generation efficiency and reliability of the wind turbine.
[0003] However, in actual operation, wind turbine blades are affected by various external factors such as wind load, temperature changes, and pressure variations, which may lead to damage to the blade structure and a decline in its health. To ensure the safe operation of wind turbine blades and extend their lifespan, health monitoring of wind turbine blades becomes crucial. Current wind turbine blade monitoring methods mainly rely on manual inspections and periodic maintenance, which suffer from low monitoring frequency, high cost, and low efficiency. Therefore, introducing advanced monitoring technologies and methods is imperative.
[0004] As can be seen from the above description, how to achieve real-time monitoring of leaf health status and improve monitoring efficiency and accuracy is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0005] To overcome the shortcomings of existing leaf monitoring methods, such as insufficient real-time performance, low efficiency, and poor accuracy, this invention proposes a method, device, medium, and equipment for monitoring the health status of leaves.
[0006] To achieve the above objectives, according to a first aspect of the present invention, an embodiment of the present invention provides a method for monitoring leaf health status, the method comprising the following steps:
[0007] Based on historical state data of the blades, a blade health status dataset is constructed, which includes load, acceleration, and temperature.
[0008] Based on the aforementioned leaf health status dataset, supervised training and unsupervised training are performed respectively to generate supervised and unsupervised models.
[0009] Real-time status data of the blades are periodically collected, and the supervised model and the unsupervised model are periodically updated.
[0010] The updated supervised model is used to monitor the health status of the leaves in real time.
[0011] Optionally, the periodic updating of the supervised model and the unsupervised model includes:
[0012] The real-time state data for a single period is input into the supervised model and the unsupervised model respectively to obtain the output results of the supervised model and the unsupervised model;
[0013] From the outputs of the supervised model and the unsupervised model, real-time state data with consistent outputs are selected and added to the leaf health state dataset to form a new leaf health state dataset.
[0014] The supervised model and the unsupervised model are updated using the new leaf health status dataset.
[0015] Optionally, the step of adding the selected real-time status data to the leaf health status dataset to form a new leaf health status dataset includes:
[0016] When the selected real-time status data is added to the leaf health status dataset, the earliest collected historical status data in the leaf health status dataset that has the same amount of data as the selected real-time status data is deleted.
[0017] Optionally, the supervised model is a deep learning extreme learning machine model based on the particle swarm optimization algorithm.
[0018] Optionally, the generated supervised model utilizes a particle swarm optimization algorithm to optimize the parameters of a deep learning extreme learning machine, including:
[0019] Initialize the parameters of the particle swarm optimization algorithm;
[0020] Initialize the parameters of the deep learning extreme learning machine;
[0021] Calculate the error of the deep learning extreme learning machine. If the error meets the first set requirement, stop training; otherwise, use the error as the fitness value of the particle swarm optimization algorithm.
[0022] Update particle position and velocity, calculate new fitness values, and update the current optimal solution;
[0023] Determine whether the second setting requirement is met. If it is met, obtain the optimal parameters of the deep learning extreme learning machine. Otherwise, update the particle position and velocity and iterate again.
[0024] Optionally, generating unsupervised models includes:
[0025] Establish state datasets for load, acceleration, and temperature respectively;
[0026] For each state dataset, a hidden Markov model is built separately.
[0027] All hidden Markov models are combined to form a total hidden Markov model, which is then used as an unsupervised model.
[0028] According to a second aspect of the present invention, embodiments of the present invention also provide a leaf health status monitoring device, comprising:
[0029] A status data acquisition unit is used to construct a blade health status dataset based on the historical status data of the blade, wherein the blade health status dataset includes load, acceleration, and temperature;
[0030] The model training unit is used to perform supervised training and unsupervised training based on the leaf health status dataset to generate supervised models and unsupervised models respectively.
[0031] The model update unit is used to periodically collect real-time status data of the blades and periodically update the supervised model and the unsupervised model.
[0032] The status monitoring unit is used to monitor the health status of the leaves in real time using the updated supervised model.
[0033] Optionally, the status data acquisition unit includes:
[0034] An optical fiber load sensor is installed at the root of the blade to collect load data at the blade root.
[0035] A fiber optic accelerometer, positioned one-third of the blade length from the blade root, is used to acquire blade vibration data.
[0036] An optical fiber temperature sensor is installed at the root of the blade to collect temperature data at the blade root.
[0037] According to a third aspect of the present invention, embodiments of the present invention also provide a computer-readable storage medium storing at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the steps of the leaf health status monitoring method in any of the above embodiments.
[0038] According to a fourth aspect of the present invention, embodiments of the present invention also provide an electronic device, including one or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and configured to be executed by the one or more processors, and the one or more programs include instructions for performing the leaf health status monitoring method of any of the above embodiments.
[0039] As described above, the leaf health status monitoring method, apparatus, medium, and device provided by the embodiments of the present invention have the following beneficial effects: A leaf health status dataset is constructed based on historical leaf state data, wherein the leaf health status dataset includes load, acceleration, and temperature; based on the leaf health status dataset, supervised training and unsupervised training are performed respectively to generate a supervised model and an unsupervised model; real-time leaf state data is periodically collected, and the supervised model and the unsupervised model are periodically updated; the updated supervised model is used to monitor the leaf health status in real time. The present invention continuously and dynamically updates the supervised learning model online by combining supervised and unsupervised learning, effectively improving the online evolution capability of the supervised learning model, thereby achieving real-time and accurate leaf health status monitoring, while eliminating the need for manual inspection and achieving high monitoring efficiency. Attached Figure Description
[0040] Figure 1 This is a flowchart illustrating a leaf health status monitoring method provided in an embodiment of the present invention;
[0041] Figure 2 This is a schematic diagram of the sensor layout provided in an embodiment of the present invention;
[0042] Figure 3 This is a flowchart illustrating a supervised model construction method provided in an embodiment of the present invention;
[0043] Figure 4 This is a flowchart illustrating an unsupervised model construction method provided in an embodiment of the present invention;
[0044] Figure 5 This is a flowchart illustrating a model update method provided in an embodiment of the present invention;
[0045] Figure 6 This is a schematic diagram of the structure of a leaf health status monitoring device provided in an embodiment of the present invention;
[0046] Figure 7 This is a schematic diagram of the hardware structure of the electronic device for performing the blade health status monitoring method provided in an embodiment of the present invention. Detailed Implementation
[0047] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this invention.
[0048] Please see Figures 1 to 7 It should be noted that the illustrations provided in this embodiment are only schematic representations of the basic concept of the present invention. Therefore, the illustrations only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.
[0049] See Figure 1 This is a flowchart illustrating a leaf health status monitoring method provided in an embodiment of the present invention, as shown below. Figure 1 As shown in the figure, the embodiment of the present invention illustrates the flow of a method for monitoring the health status of leaves.
[0050] Step S101: Construct a leaf health status dataset based on the historical state data of the leaves.
[0051] The blade health status dataset includes load, acceleration, and temperature.
[0052] In practical implementation, fiber optic sensor technology, as a novel sensor technology, offers numerous advantages in structural monitoring, such as high sensitivity, resistance to electromagnetic interference, and ease of deployment. Fiber optic sensors can monitor parameters such as strain and vibration of wind turbine blades in real time, providing crucial information on the health status of the blade structure. On the other hand, with the rapid development of artificial intelligence and machine learning, machine learning methods have also demonstrated enormous potential in the field of structural health monitoring. Machine learning can analyze large amounts of monitoring data, identify potential problems, and predict blade lifespan, providing a scientific basis for the maintenance and management of wind turbine blades.
[0053] To ensure the accuracy of data acquisition and thus improve the precision of leaf health status monitoring, in an exemplary embodiment, see [reference needed]. Figure 2 This is a schematic diagram of the sensor layout provided in an embodiment of the present invention, as shown below. Figure 2 As shown, for each wind turbine blade 1, four fiber optic load sensors 2 and three bidirectional MEMS fiber optic accelerometers 3 can be installed at the blade root and one-third of the distance from the blade root of each blade 1. Furthermore, in practical implementation, a fiber optic vibration sensor (not shown) can also be installed next to each fiber optic load sensor to perform real-time sampling at 100Hz, enabling second-level threshold alarms or early warnings for load and amplitude. Figure 2 The sensor layout shown includes an optical fiber load sensor 2 located at the blade root for collecting load data at the blade root; an optical fiber accelerometer 3 located at 1 / 3 of the blade length from the blade root for collecting blade vibration data; and an optical fiber temperature sensor 4 located at the blade root for collecting temperature data at the blade root.
[0054] Step S102: Based on the leaf health status dataset, supervised training and unsupervised training are performed respectively to generate supervised models and unsupervised models.
[0055] In this embodiment of the invention, the supervised model is a deep learning extreme learning machine model based on the particle swarm optimization algorithm. Particle Swarm Optimization (PSO) is an adaptive optimization algorithm that simulates the information sharing and cooperation strategies in the foraging behavior of bird flocks. This algorithm searches for the optimal solution to the problem through iterative search. The PSO algorithm includes two key steps: updating the velocity and position of the particles, and iteratively updating the particle positions so that the entire particle swarm gradually converges to the optimal solution.
[0056] Deep Extreme Learning Machine (DELM) is a single-hidden-layer feedforward neural network characterized by randomly initializing the connection weights between the input layer and the hidden layer, and then using the least squares method to optimize the output weights of the hidden layer. DELM achieves feature extraction and data modeling by randomly selecting and training hidden layer nodes.
[0057] Specifically, in order to improve the performance of the Deep Learning Extreme Learning Machine (DELM), this application combines the Particle Swarm Optimization (PSO) algorithm with the Deep Learning Extreme Learning Machine (DELM) and uses the PSO algorithm to optimize the parameters of the Deep Learning Extreme Learning Machine (DELM).
[0058] See Figure 3 Figure 3 is a flowchart illustrating a supervised model construction method provided in an embodiment of the present invention. As shown in Figure 3, the method includes the following steps:
[0059] Step S1021: Initialize the parameters of the particle swarm optimization algorithm.
[0060] Step S1022: Initialize the parameters of the deep learning extreme learning machine.
[0061] Step S1023: Calculate the error of the deep learning extreme learning machine. If the error meets the first set requirement, stop training; otherwise, use the error as the fitness value of the particle swarm optimization algorithm.
[0062] Step S1024: Update particle position and velocity, calculate new fitness values, and update the current optimal solution.
[0063] Step S1025: Determine whether the second setting requirement is met. If it is met, obtain the optimal parameters of the deep learning extreme learning machine. Otherwise, update the particle position and velocity and iterate again.
[0064] In this embodiment of the invention, the unsupervised model is built based on the Hidden Markov Model (HMM). A Hidden Markov Model (HMM) is a statistical model used to describe Markov processes containing hidden unknown parameters. An HMM is a type of Markov chain whose states cannot be directly observed but can be observed through a sequence of observation vectors. Each observation vector represents various states through a certain probability density distribution, and each observation vector is generated by a sequence of states with a corresponding probability density distribution. Therefore, an HMM is a doubly stochastic process, consisting of a Hidden Markov chain with a certain number of states and a set of explicit random functions.
[0065] See Figure 4 This is a flowchart illustrating an unsupervised model construction method provided in an embodiment of the present invention, as shown below. Figure 4 As shown, the method includes the following steps:
[0066] Step S1026: Establish state datasets for load, acceleration, and temperature respectively.
[0067] Step S1027: Based on each state dataset, establish a Hidden Markov Model.
[0068] Step S1028: Combine all hidden Markov models to form a total hidden Markov model, and use the total hidden Markov model as an unsupervised model.
[0069] Step S103: Periodically collect real-time status data of the blades and periodically update the supervised model and the unsupervised model.
[0070] To achieve real-time updates of both supervised and unsupervised models, see [link / reference]. Figure 5 This is a flowchart illustrating a model update method provided in an embodiment of the present invention, as shown below. Figure 5 As shown, this method illustrates the model update process:
[0071] Step S1031: Input the real-time state data of a single period into the supervised model and the unsupervised model respectively to obtain the output results of the supervised model and the unsupervised model.
[0072] Step S1032: From the output results of the supervised model and the unsupervised model, select real-time state data with consistent output results, and add the selected real-time state data to the leaf health state dataset to form a new leaf health state dataset.
[0073] In practice, when the selected real-time status data is added to the leaf health status dataset, the earliest collected historical status data in the leaf health status dataset that has the same amount of data as the selected real-time status data is deleted.
[0074] Step S1033: Update the supervised model and the unsupervised model using the new leaf health status dataset.
[0075] Step S104: Use the updated supervised model to monitor the health status of the leaves in real time.
[0076] As described in the above embodiments, the leaf health status monitoring method provided by this invention constructs a leaf health status dataset based on historical leaf status data, wherein the leaf health status dataset includes load, acceleration, and temperature; based on the leaf health status dataset, supervised training and unsupervised training are performed respectively to generate supervised and unsupervised models; real-time leaf status data is periodically collected, and the supervised and unsupervised models are periodically updated; the updated supervised model is used to monitor the leaf health status in real time. This invention continuously and dynamically updates the supervised learning model online by combining supervised and unsupervised learning, effectively improving the online evolution capability of the supervised learning model, thereby achieving real-time and accurate leaf health status monitoring, while eliminating the need for manual inspection and achieving high monitoring efficiency. Moreover, through the training method combining unsupervised and supervised learning, unsupervised learning can automatically identify leaf health status without labels, effectively solving the problem of insufficient labels for leaf health status data, and greatly improving the accuracy of online leaf monitoring.
[0077] Through the description of the above method embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0078] This invention provides a non-volatile computer storage medium storing computer-executable instructions that can execute the leaf health status monitoring method in any of the above method embodiments.
[0079] Corresponding to the leaf health status monitoring method embodiment provided by the present invention, the present invention also provides a leaf health status monitoring device.
[0080] See Figure 6 Figure 1 is a schematic diagram of a leaf health status monitoring device provided in an embodiment of the present invention. As shown in the figure, the device includes:
[0081] The status data acquisition unit 11 is used to construct a blade health status dataset based on the historical status data of the blade, wherein the blade health status dataset includes load, acceleration, and temperature;
[0082] The model training unit 12 is used to perform supervised training and unsupervised training based on the leaf health status dataset to generate supervised and unsupervised models.
[0083] The model update unit 13 is used to periodically collect real-time status data of the blades and periodically update the supervised model and the unsupervised model.
[0084] The status monitoring unit 14 is used to monitor the health status of the leaves in real time using the updated supervised model.
[0085] Optionally, see Figure 2 The status data acquisition unit 11 includes:
[0086] The fiber optic load sensor 2, installed at the root of the blade, is used to collect load data at the root of the blade.
[0087] An optical fiber accelerometer, 3, is installed at a distance of 1 / 3 of the blade length from the blade root to collect blade vibration data.
[0088] The fiber optic temperature sensor 4, installed at the root of the blade, is used to collect temperature data at the root of the blade.
[0089] For the specific sensor layout structure, please refer to the description of the above embodiments, which will not be repeated here.
[0090] Optionally, the model update unit 13 is further configured to:
[0091] The real-time state data for a single period is input into the supervised model and the unsupervised model respectively to obtain the output results of the supervised model and the unsupervised model;
[0092] From the outputs of the supervised model and the unsupervised model, real-time state data with consistent outputs are selected and added to the leaf health state dataset to form a new leaf health state dataset.
[0093] The supervised model and the unsupervised model are updated using the new leaf health status dataset.
[0094] In an exemplary embodiment, the model update unit 13 is further configured to delete the earliest collected historical status data in the leaf health status dataset that has the same amount of data as the selected real-time status data when the selected real-time status data is added to the leaf health status dataset.
[0095] Optionally, the supervised model is a deep learning extreme learning machine model based on the particle swarm optimization algorithm. The model training unit 12 uses the particle swarm optimization algorithm to optimize the parameters of the deep learning extreme learning machine, including:
[0096] Initialize the parameters of the particle swarm optimization algorithm;
[0097] Initialize the parameters of the deep learning extreme learning machine;
[0098] Calculate the error of the deep learning extreme learning machine. If the error meets the first set requirement, stop training; otherwise, use the error as the fitness value of the particle swarm optimization algorithm.
[0099] Update particle position and velocity, calculate new fitness values, and update the current optimal solution;
[0100] Determine whether the second setting requirement is met. If it is met, obtain the optimal parameters of the deep learning extreme learning machine. Otherwise, update the particle position and velocity and iterate again.
[0101] Optionally, the model training unit 12 is further configured to generate an unsupervised model, including:
[0102] Establish state datasets for load, acceleration, and temperature respectively;
[0103] For each state dataset, a hidden Markov model is built separately.
[0104] All hidden Markov models are combined to form a total hidden Markov model, which is then used as an unsupervised model.
[0105] Figure 7 This is a schematic diagram of the hardware structure of the electronic device for performing the blade health status monitoring method provided in an embodiment of the present invention, as shown below. Figure 7 As shown, the device includes:
[0106] One or more processors 710 and memory 720, Figure 7 Taking the 710 processor as an example, Figure 2 The fiber optic load sensor 2, fiber optic accelerometer 3, and fiber optic temperature sensor 4 shown are all connected to the processor 710 for communication, and are used to transmit the collected load, acceleration, temperature and other data of the blade to the processor 710 for data processing.
[0107] The device for performing the blade health status monitoring method may also include: an input device 730 and an output device 740.
[0108] The processor 710, memory 720, input device 730, and output device 740 can be connected via a bus or other means. Figure 7 Taking the example of a connection between China and Israel via a bus.
[0109] The memory 720, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the battery health state estimation method in this embodiment of the invention (e.g., attached...). Figure 6 The data acquisition unit 11, model training unit 12, model update unit 13, and status monitoring unit 14 are shown. The processor 710 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 720, thereby realizing the leaf health status monitoring method of the above-described method embodiment.
[0110] The memory 720 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the blade health monitoring device. Furthermore, the memory 720 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 720 may optionally include memory remotely located relative to the processor 710, and these remote memories can be connected to the blade health monitoring device via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.
[0111] The input device 730 can receive input digital or character information, and generate key signal inputs related to user settings and function control of the blade health status monitoring device. The output device 740 may include a display screen or other display device.
[0112] The one or more modules are stored in the memory 720, and when executed by the one or more processors 710, they execute the leaf health status monitoring method in any of the above method embodiments.
[0113] The above-described product can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.
[0114] The electronic devices of this invention exist in various forms, including but not limited to:
[0115] (1) Mobile communication devices: These devices are characterized by their mobile communication capabilities and primarily aim to provide voice and data communication. These terminals include: smartphones (e.g., iPhones), multimedia phones, feature phones, and low-end phones, etc.
[0116] (2) Ultra-mobile personal computer devices: These devices fall under the category of personal computers, possessing computing and processing capabilities, and generally also have mobile internet access features. These terminals include PDAs, MIDs, and UMPCs, such as the iPad.
[0117] (3) Portable entertainment devices: These devices can display and play multimedia content. This category includes: audio and video players (such as iPods), handheld game consoles, e-books, as well as smart toys and portable car navigation devices.
[0118] (4) Server: A device that provides computing services. The components of a server include a processor, hard disk, memory, system bus, etc. Servers are similar to general computer architectures, but because they need to provide highly reliable services, they have higher requirements in terms of processing power, stability, reliability, security, scalability, and manageability.
[0119] (5) Other electronic devices with data interaction functions.
[0120] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0121] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.
[0122] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0123] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for monitoring leaf health status, characterized in that, include: Based on historical state data of the blades, a blade health status dataset is constructed, which includes load, acceleration, and temperature. Based on the aforementioned leaf health status dataset, supervised training and unsupervised training are performed respectively to generate supervised and unsupervised models. Real-time status data of the blades are periodically collected, and the supervised model and the unsupervised model are periodically updated. The updated supervised model is used to monitor the health status of the leaves in real time.
2. The leaf health status monitoring method according to claim 1, characterized in that, The periodic updating of the supervised model and the unsupervised model includes: The real-time state data for a single period is input into the supervised model and the unsupervised model respectively to obtain the output results of the supervised model and the unsupervised model; From the outputs of the supervised model and the unsupervised model, real-time state data with consistent outputs are selected and added to the leaf health state dataset to form a new leaf health state dataset. The supervised model and the unsupervised model are updated using the new leaf health status dataset.
3. The leaf health status monitoring method according to claim 2, characterized in that, The step of adding the selected real-time status data to the leaf health status dataset to form a new leaf health status dataset includes: When the selected real-time status data is added to the leaf health status dataset, the earliest collected historical status data in the leaf health status dataset that has the same amount of data as the selected real-time status data is deleted.
4. The leaf health status monitoring method according to any one of claims 1 to 3, characterized in that, The supervised model is a deep learning extreme learning machine model based on the particle swarm optimization algorithm.
5. The leaf health status monitoring method according to claim 4, characterized in that, The generated supervised model utilizes the particle swarm optimization algorithm to optimize the parameters of the deep learning extreme learning machine, including: Initialize the parameters of the particle swarm optimization algorithm; Initialize the parameters of the deep learning extreme learning machine; Calculate the error of the deep learning extreme learning machine. If the error meets the first set requirement, stop training; otherwise, use the error as the fitness value of the particle swarm optimization algorithm. Update particle position and velocity, calculate new fitness values, and update the current optimal solution; Determine whether the second setting requirement is met. If it is met, obtain the optimal parameters of the deep learning extreme learning machine. Otherwise, update the particle position and velocity and iterate again.
6. The leaf health status monitoring method according to any one of claims 1 to 3, characterized in that, Generative unsupervised models include: Establish state datasets for load, acceleration, and temperature respectively; For each state dataset, a hidden Markov model is built separately. All hidden Markov models are combined to form a total hidden Markov model, which is then used as an unsupervised model.
7. A leaf health status monitoring device, characterized in that, include: A status data acquisition unit is used to construct a blade health status dataset based on the historical status data of the blade, wherein the blade health status dataset includes load, acceleration, and temperature; The model training unit is used to perform supervised training and unsupervised training based on the leaf health status dataset to generate supervised models and unsupervised models respectively. The model update unit is used to periodically collect real-time status data of the blades and periodically update the supervised model and the unsupervised model. The status monitoring unit is used to monitor the health status of the leaves in real time using the updated supervised model.
8. The leaf health status monitoring device according to claim 7, characterized in that, The status data acquisition unit includes: An optical fiber load sensor is installed at the root of the blade to collect load data at the blade root. A fiber optic accelerometer, positioned one-third of the blade length from the blade root, is used to acquire blade vibration data. An optical fiber temperature sensor is installed at the root of the blade to collect temperature data at the blade root.
9. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the steps of the blade health status monitoring method as described in any one of claims 1 to 6.
10. An electronic device, characterized in that, It includes one or more processors, one or more memories, and one or more programs, wherein the one or more programs are stored in the one or more memories and configured to be executed by the one or more processors, and the one or more programs include instructions for performing the blade health status monitoring method as described in any one of claims 1-6.